ble variations which have
arisen independently of the nature of the conditions. If, for instance, a
plant had to be modified so as to become fitted to inhabit a humid instead
of an arid station, we have no reason to believe that variations of the
right kind would occur more frequently if the parent-plant inhabited a
station a little more {291} humid than usual. Whether the station was
unusually dry or humid, variations adapting the plant in a slight degree
for directly opposite habits of life would occasionally arise, as we have
reason to believe from what we know in other cases.

In most, perhaps in all cases, the organisation or constitution of the
being which is acted on, is a much more important element than the nature
of the changed conditions, in determining the nature of the variation. We
have evidence of this in the appearance of nearly similar modifications
under different conditions, and of different modifications under apparently
nearly the same conditions. We have still better evidence of this in
closely parallel varieties being frequently produced from distinct races,
or even distinct species, and in the frequent recurrence of the same
monstrosity in the same species. We have also seen that the degree to which
domesticated birds have varied, does not stand in any close relation with
the amount of change to which they have been subjected.

To recur once again to bud-variations. When we reflect on the millions of
buds which many trees have produced, before some one bud has varied, we are
lost in wonder what the precise cause of each variation can be. Let us
recall the case given by Andrew Knight of the forty-year-old tree of the
yellow magnum bonum plum, an old variety which has been propagated by
grafts on various stocks for a very long period throughout Europe and North
America, and on which a single bud suddenly produced the red magnum bonum.
We should also bear in mind that distinct varieties, and even distinct
species,--as in the case of peaches, nectarines, and apricots,--of certain
roses and camellias,--although separated by a vast number of generations
from any progenitor in common, and although cultivated under diversified
conditions, have yielded by bud-variation closely analogous varieties. When
we reflect on these facts we become deeply impressed with the conviction
that in such cases the nature of the variation depends but little on the
conditions to which the plant has been exposed, and not in an